Intelligent garbage collection based on content similarity
AssigneePure Storage, Inc.
InventorsETHAN L. MILLER, JOHN COLGROVE, RONALD KARR, ROBERT LEE
Life of this patent
Estimating the term…
Abstract
A storage system performs garbage collection, with data compression, in storage memory. The system obtains hash results from data segments. The system determines similarity of content of data segments, based on the hash results. The system performs data compression of live data of two or more data segments that have similarity of content meeting a similarity threshold. The system writes the compressed live data of the two or more data segments into the storage memory.
Claims20
A storage system, comprising:
- storage memory;
- and
- a processing device operatively coupled to the storage memory, the processing device configured to:
- determine similarity of content of a plurality of data segments stored in the storage memory based on a plurality of hash values associated with the plurality of data segments, the plurality of data segments comprising live data and dead data;
- perform garbage collection of the dead data for the plurality of data segments, wherein performing the garbage collection comprises compressing the live data based on a similarity threshold;
- and
- write the compressed live data into the storage memory.
A method, comprising:
- determining similarity of content of a plurality of data segments stored in a storage memory of a storage system based on a plurality of hash values associated with the plurality of data segments, the plurality of data segments comprising live data and dead data;
- performing garbage collection of the dead data for the plurality of data segments, wherein performing the garbage collection comprises compressing the live data based on a similarity threshold;
- and
- writing the compressed live data into the storage memory.
A non-transitory computer-readable media having instructions thereupon which, when executed by a processing device, cause the processing device to:
- determine similarity of content of a plurality of data segments stored in a storage memory of a storage system based on a plurality of hash values associated with the plurality of data segments, the plurality of data segments comprising live data and dead data;
- perform garbage collection of the dead data for the plurality of data segments, wherein performing the garbage collection comprises compressing the live data based on a similarity threshold;
- and
- write the compressed live data into the storage memory.
Dependencies are inferred from the claim text.